AI law to be voted on in Europe - BBC News

Demo Transcription

Transcription Details:

Date: Jun 16, 2023
Type: YouTube
Source: English
Destinations: zh-CN, fr, it, es
Duration: 0:02:11

AI law to be voted on in Europe - BBC News

Multiple sepakers transcription text:

SPEAKER 1: Politicians in Europe will vote on a proposal to bring in a law that would govern the use of artificial intelligence. The move comes after warnings from experts of serious risks and consequences if it continues to go unregulated. Here's Simi Jolla-Osso with more.

SPEAKER 2: Artificial intelligence has been in the news a lot lately, whether it's students using apps like ChatGPT to write up essays, or medical experts using AI tools to advance research on cancer. But then there are also those AI -generated photos, videos and music making the rounds, and those have sparked concerns around copyright and also around the spread of disinformation. So lawmakers, specifically in Europe, have decided to do something about it, introducing the Artificial Intelligence Act. Members of the European Parliament want to make sure AI is safe, transparent, traceable, non -discriminatory and environmentally friendly. And if approved, it will be the first of its kind in the world. The idea is to govern the use of AI based on three levels of risk, starting with unacceptable risk. That is, when AI is used for things deemed so unethical, such as biometric surveillance, or even using it to keep social scores on people. Think Netflix's Black Mirror. Next is high risk, so things that might cause harm to people's health, harm to the environment, or affect people's fundamental rights. For example, an AI tool that scans CVs in order to rank job applicants, which is fine, as long as it abides by certain rules. For example, it doesn't discriminate against applicants based on their race, age or gender. Next are AI apps and tools that aren't banned and don't pose any high risk. So when will this AI act become law? Well, it might still take some time. The European Parliament are set to vote on the current draft in June, and even then it will only become law once both the European Parliament and the European Council have agreed on it. Thanks for watching.

Transcription subtitles:

0
00:00:00,740 --> 00:00:06,440
Politicians in Europe will vote on a proposal to bring in a law that would govern the use of artificial intelligence.

1
00:00:06,860 --> 00:00:13,080
The move comes after warnings from experts of serious risks and consequences if it continues to go unregulated.

2
00:00:13,660 --> 00:00:15,960
Here's Simi Jolla-Osso with more.

3
00:00:17,340 --> 00:00:20,900
Artificial intelligence has been in the news a lot lately,

4
00:00:21,040 --> 00:00:24,880
whether it's students using apps like ChatGPT to write up essays,

5
00:00:25,000 --> 00:00:29,980
or medical experts using AI tools to advance research on cancer.

6
00:00:30,360 --> 00:00:33,260
But then there are also those AI-generated photos,

7
00:00:33,400 --> 00:00:35,340
videos and music making the rounds,

8
00:00:36,000 --> 00:00:41,580
and those have sparked concerns around copyright and also around the spread of disinformation.

9
00:00:41,960 --> 00:00:42,500
So lawmakers,

10
00:00:43,040 --> 00:00:43,900
specifically in Europe,

11
00:00:44,340 --> 00:00:46,440
have decided to do something about it,

12
00:00:46,660 --> 00:00:49,520
introducing the Artificial Intelligence Act.

13
00:00:49,560 --> 00:00:53,800
Members of the European Parliament want to make sure AI is safe,

14
00:00:54,320 --> 00:00:54,860
transparent,

15
00:00:55,300 --> 00:00:55,940
traceable,

16
00:00:56,260 --> 00:00:59,080
non-discriminatory and environmentally friendly.

17
00:00:59,320 --> 00:01:01,820
And if approved, it will be the first of its

18
00:01:01,820 --> 00:01:03,140
kind in the world.

19
00:01:03,580 --> 00:01:07,900
The idea is to govern the use of AI based on three levels of risk,

20
00:01:08,200 --> 00:01:10,600
starting with unacceptable risk.

21
00:01:11,080 --> 00:01:14,980
That is, when AI is used for things deemed so unethical,

22
00:01:15,100 --> 00:01:16,920
such as biometric surveillance,

23
00:01:17,620 --> 00:01:21,160
or even using it to keep social scores on people.

24
00:01:21,440 --> 00:01:23,120
Think Netflix's Black Mirror.

25
00:01:23,600 --> 00:01:24,960
Next is high risk,

26
00:01:25,080 --> 00:01:28,300
so things that might cause harm to people's health,

27
00:01:28,740 --> 00:01:29,720
harm to the environment,

28
00:01:29,920 --> 00:01:32,140
or affect people's fundamental rights.

29
00:01:32,660 --> 00:01:37,220
For example, an AI tool that scans CVs in order to rank job applicants,

30
00:01:37,260 --> 00:01:40,480
which is fine, as long as it abides by certain

31
00:01:40,480 --> 00:01:40,880
rules.

32
00:01:41,000 --> 00:01:46,060
For example, it doesn't discriminate against applicants based on their race, age or gender.

33
00:01:46,600 --> 00:01:53,120
Next are AI apps and tools that aren't banned and don't pose any high risk.

34
00:01:53,300 --> 00:01:56,340
So when will this AI act become law?

35
00:01:56,580 --> 00:01:57,040
Well,

36
00:01:57,400 --> 00:01:58,760
it might still take some time.

37
00:01:58,940 --> 00:02:02,920
The European Parliament are set to vote on the current draft in June,

38
00:02:02,920 --> 00:02:04,420
and even then it will only

39
00:02:04,420 --> 00:02:09,900
become law once both the European Parliament and the European Council have agreed on it.

40
00:02:10,480 --> 00:02:10,560
Thanks for watching.

Translations:

Chinese (Simplified) transcription text:

Chinese (Simplified) transcription subtitles:

0
00:00:00,740 --> 00:00:06,440
欧洲的政客们将对一项提案进行投票，以制定一项管理人工智能使用的法律。

1
00:00:06,860 --> 00:00:13,080
此举是在专家警告称，如果继续不受监管，将面临严重风险和后果。

2
00:00:13,660 --> 00:00:15,960
这是 Simi Jolla-Osso 的更多内容。

3
00:00:17,340 --> 00:00:20,900
人工智能最近经常出现在新闻中，

4
00:00:21,040 --> 00:00:24,880
无论是学生使用 ChatGPT 等应用来撰写论文，

5
00:00:25,000 --> 00:00:29,980
或使用人工智能工具推进癌症研究的医学专家。

6
00:00:30,360 --> 00:00:33,260
但还有那些人工智能生成的照片，

7
00:00:33,400 --> 00:00:35,340
流传的视频和音乐，

8
00:00:36,000 --> 00:00:41,580
这些引发了对版权和虚假信息传播的担忧。

9
00:00:41,960 --> 00:00:42,500
所以立法者，

10
00:00:43,040 --> 00:00:43,900
特别是在欧洲，

11
00:00:44,340 --> 00:00:46,440
决定为此做点什么，

12
00:00:46,660 --> 00:00:49,520
引入人工智能法案。

13
00:00:49,560 --> 00:00:53,800
欧洲议会议员希望确保人工智能是安全的，

14
00:00:54,320 --> 00:00:54,860
透明的，

15
00:00:55,300 --> 00:00:55,940
可追溯，

16
00:00:56,260 --> 00:00:59,080
非歧视和环保。

17
00:00:59,320 --> 00:01:01,820
如果获得批准，这将是其第一个

18
00:01:01,820 --> 00:01:03,140
善良的世界。

19
00:01:03,580 --> 00:01:07,900
这个想法是根据三个风险级别来管理人工智能的使用，

20
00:01:08,200 --> 00:01:10,600
从不可接受的风险开始。

21
00:01:11,080 --> 00:01:14,980
也就是说，当人工智能被用于被认为不道德的事情时，

22
00:01:15,100 --> 00:01:16,920
例如生物识别监控，

23
00:01:17,620 --> 00:01:21,160
甚至用它来保持人们的社会分数。

24
00:01:21,440 --> 00:01:23,120
想想 Netflix 的黑镜。

25
00:01:23,600 --> 00:01:24,960
接下来是高风险，

26
00:01:25,080 --> 00:01:28,300
所以可能对人的健康造成危害的事情，

27
00:01:28,740 --> 00:01:29,720
对环境的危害，

28
00:01:29,920 --> 00:01:32,140
或影响人们的基本权利。

29
00:01:32,660 --> 00:01:37,220
例如，扫描简历以对求职者进行排名的 AI 工具，

30
00:01:37,260 --> 00:01:40,480
这很好，只要它遵守某些

31
00:01:40,480 --> 00:01:40,880
规则。

32
00:01:41,000 --> 00:01:46,060
例如，它不会根据种族、年龄或性别歧视申请人。

33
00:01:46,600 --> 00:01:53,120
接下来是未被禁止且不会构成任何高风险的 AI 应用程序和工具。

34
00:01:53,300 --> 00:01:56,340
那么这项人工智能法案何时会成为法律呢？

35
00:01:56,580 --> 00:01:57,040
出色地，

36
00:01:57,400 --> 00:01:58,760
可能还需要一些时间。

37
00:01:58,940 --> 00:02:02,920
欧洲议会将于 6 月就目前的草案进行投票，

38
00:02:02,920 --> 00:02:04,420
即便如此，它也只会

39
00:02:04,420 --> 00:02:09,900
一旦欧洲议会和欧洲理事会就此达成一致，就成为法律。

40
00:02:10,480 --> 00:02:10,560
感谢收看。

French transcription text:

INTERVENANT 1 : Les politiciens européens voteront sur une proposition visant à faire adopter une loi qui régirait l'utilisation de l'intelligence artificielle. Cette décision intervient après des avertissements d'experts sur les risques graves et les conséquences si elle continue à ne pas être réglementée. Voici Simi Jolla -Osso avec plus.

INTERVENANT 2: L'intelligence artificielle a beaucoup fait les manchettes ces derniers temps, qu'il s'agisse d'étudiants utilisant des applications comme ChatGPT pour rédiger des essais, ou d'experts médicaux utilisant des outils d'IA pour faire avancer la recherche sur cancer. Mais il y a aussi ces photos, vidéos et musiques générées par l'IA qui circulent, et celles-ci ont suscité des inquiétudes concernant le droit d'auteur et la propagation de la désinformation. Les législateurs, en particulier en Europe, ont donc décidé de faire quelque chose à ce sujet, en introduisant la loi sur l'intelligence artificielle. Les membres du Parlement européen veulent s'assurer que l'IA est sûre, transparente, traçable, non discriminatoire et respectueuse de l'environnement. Et s'il est approuvé, ce sera le premier du genre au monde. L'idée est de régir l'utilisation de l'IA en fonction de trois niveaux de risque, en commençant par le risque inacceptable. C'est-à-dire lorsque l'IA est utilisée pour des choses jugées si contraires à l'éthique, comme la surveillance biométrique, ou même l'utiliser pour garder des scores sociaux sur les gens. Pensez au miroir noir de Netflix. Vient ensuite le risque élevé, c'est-à-dire les choses qui pourraient nuire à la santé des personnes, nuire à l'environnement ou affecter les droits fondamentaux des personnes. Par exemple, un outil d'intelligence artificielle qui scanne les CV afin de classer les candidats à un emploi, ce qui est bien, tant qu'il respecte certaines règles. Par exemple, il ne discrimine pas les candidats en fonction de leur race, de leur âge ou de leur sexe. Viennent ensuite les applications et les outils d'IA qui ne sont pas interdits et ne présentent aucun risque élevé. Alors, quand cette loi sur l'IA deviendra-t-elle loi ? Eh bien, cela peut encore prendre du temps. Le Parlement européen devrait voter sur le projet actuel en juin, et même alors, il ne deviendra loi qu'une fois que le Parlement européen et le Conseil européen se seront mis d'accord dessus. Merci d'avoir regardé.

French transcription subtitles:

0
00:00:00,740 --> 00:00:06,440
Les politiciens européens voteront sur une proposition visant à faire adopter une loi qui régirait l'utilisation de l'intelligence artificielle.

1
00:00:06,860 --> 00:00:13,080
Cette décision intervient après des avertissements d'experts sur les risques graves et les conséquences si elle continue à ne pas être réglementée.

2
00:00:13,660 --> 00:00:15,960
Voici Simi Jolla-Osso avec plus.

3
00:00:17,340 --> 00:00:20,900
L'intelligence artificielle a fait beaucoup parler d'elle ces derniers temps,

4
00:00:21,040 --> 00:00:24,880
qu'il s'agisse d'étudiants utilisant des applications comme ChatGPT pour rédiger des dissertations,

5
00:00:25,000 --> 00:00:29,980
ou des experts médicaux utilisant des outils d'IA pour faire avancer la recherche sur le cancer.

6
00:00:30,360 --> 00:00:33,260
Mais il y a aussi ces photos générées par l'IA,

7
00:00:33,400 --> 00:00:35,340
des vidéos et de la musique font le tour,

8
00:00:36,000 --> 00:00:41,580
et ceux-ci ont suscité des inquiétudes concernant le droit d'auteur et également autour de la propagation de la désinformation.

9
00:00:41,960 --> 00:00:42,500
Alors les législateurs,

10
00:00:43,040 --> 00:00:43,900
spécifiquement en Europe,

11
00:00:44,340 --> 00:00:46,440
ont décidé de faire quelque chose à ce sujet,

12
00:00:46,660 --> 00:00:49,520
l'introduction de la loi sur l'intelligence artificielle.

13
00:00:49,560 --> 00:00:53,800
Les membres du Parlement européen veulent s'assurer que l'IA est sûre,

14
00:00:54,320 --> 00:00:54,860
transparent,

15
00:00:55,300 --> 00:00:55,940
traçable,

16
00:00:56,260 --> 00:00:59,080
non discriminatoire et respectueux de l'environnement.

17
00:00:59,320 --> 00:01:01,820
Et s'il est approuvé, ce sera le premier de ses

18
00:01:01,820 --> 00:01:03,140
genre dans le monde.

19
00:01:03,580 --> 00:01:07,900
L'idée est de gouverner l'utilisation de l'IA en fonction de trois niveaux de risque,

20
00:01:08,200 --> 00:01:10,600
à commencer par un risque inacceptable.

21
00:01:11,080 --> 00:01:14,980
Autrement dit, lorsque l'IA est utilisée pour des choses jugées si contraires à l'éthique,

22
00:01:15,100 --> 00:01:16,920
comme la surveillance biométrique,

23
00:01:17,620 --> 00:01:21,160
ou même l'utiliser pour garder des scores sociaux sur les gens.

24
00:01:21,440 --> 00:01:23,120
Pensez au miroir noir de Netflix.

25
00:01:23,600 --> 00:01:24,960
Vient ensuite le risque élevé,

26
00:01:25,080 --> 00:01:28,300
donc les choses qui pourraient nuire à la santé des gens,

27
00:01:28,740 --> 00:01:29,720
atteinte à l'environnement,

28
00:01:29,920 --> 00:01:32,140
ou affecter les droits fondamentaux des personnes.

29
00:01:32,660 --> 00:01:37,220
Par exemple, un outil d'IA qui scanne les CV afin de classer les candidats à un emploi,

30
00:01:37,260 --> 00:01:40,480
ce qui est bien, tant qu'il respecte certaines

31
00:01:40,480 --> 00:01:40,880
règles.

32
00:01:41,000 --> 00:01:46,060
Par exemple, il ne discrimine pas les candidats en fonction de leur race, de leur âge ou de leur sexe.

33
00:01:46,600 --> 00:01:53,120
Viennent ensuite les applications et les outils d'IA qui ne sont pas interdits et ne présentent aucun risque élevé.

34
00:01:53,300 --> 00:01:56,340
Alors, quand cette loi sur l'IA deviendra-t-elle loi ?

35
00:01:56,580 --> 00:01:57,040
Bien,

36
00:01:57,400 --> 00:01:58,760
cela peut encore prendre du temps.

37
00:01:58,940 --> 00:02:02,920
Le Parlement européen devrait voter sur le projet actuel en juin,

38
00:02:02,920 --> 00:02:04,420
et même alors ça ne fera que

39
00:02:04,420 --> 00:02:09,900
devenir une loi une fois que le Parlement européen et le Conseil européen se seront mis d'accord dessus.

40
00:02:10,480 --> 00:02:10,560
Merci d'avoir regardé.

Italian transcription text:

RELATORE 1: I politici in Europa voteranno su una proposta per introdurre una legge che regolerebbe l'uso dell'intelligenza artificiale. La mossa arriva dopo gli avvertimenti degli esperti di gravi rischi e conseguenze se continua a non essere regolamentata. Ecco Simi Jolla -Osso con altro.

SPEAKER 2: Ultimamente l'intelligenza artificiale ha fatto molto parlare di sé, sia che si tratti di studenti che utilizzano app come ChatGPT per scrivere saggi o di esperti medici che utilizzano strumenti di intelligenza artificiale per far progredire la ricerca su cancro. Ma poi ci sono anche quelle foto, video e musica generati dall'intelligenza artificiale che fanno il giro, e quelli hanno suscitato preoccupazioni sul copyright e anche sulla diffusione della disinformazione. Così i legislatori, in particolare in Europa, hanno deciso di fare qualcosa al riguardo, introducendo l'Artificial Intelligence Act. I membri del Parlamento europeo vogliono assicurarsi che l'IA sia sicura, trasparente, tracciabile, non discriminatoria e rispettosa dell'ambiente. E se approvato, sarà il primo del suo genere al mondo. L'idea è di governare l'uso dell'IA sulla base di tre livelli di rischio, a partire dal rischio inaccettabile. Cioè, quando l'intelligenza artificiale viene utilizzata per cose ritenute così non etiche, come la sorveglianza biometrica, o addirittura per mantenere i punteggi sociali sulle persone. Pensa a Black Mirror di Netflix. Il prossimo è ad alto rischio, quindi cose che potrebbero causare danni alla salute delle persone, danni all'ambiente o ledere i diritti fondamentali delle persone. Ad esempio, uno strumento di intelligenza artificiale che scansiona i CV per classificare i candidati al lavoro, il che va bene, purché rispetti determinate regole. Ad esempio, non discrimina i candidati in base alla loro razza, età o sesso. Successivamente ci sono app e strumenti di intelligenza artificiale che non sono vietati e non presentano alcun rischio elevato. Quindi, quando questo atto di intelligenza artificiale diventerà legge? Beh, potrebbe volerci ancora del tempo. Il Parlamento europeo voterà l'attuale progetto a giugno, e anche allora diventerà legge solo una volta che sia il Parlamento europeo che il Consiglio europeo lo avranno concordato. Grazie per la visione.

Italian transcription subtitles:

0
00:00:00,740 --> 00:00:06,440
I politici in Europa voteranno su una proposta per introdurre una legge che regolerebbe l'uso dell'intelligenza artificiale.

1
00:00:06,860 --> 00:00:13,080
La mossa arriva dopo gli avvertimenti degli esperti di gravi rischi e conseguenze se continua a non essere regolamentata.

2
00:00:13,660 --> 00:00:15,960
Ecco Simi Jolla-Osso con altro.

3
00:00:17,340 --> 00:00:20,900
L'intelligenza artificiale è stata molto nelle notizie ultimamente,

4
00:00:21,040 --> 00:00:24,880
che si tratti di studenti che utilizzano app come ChatGPT per scrivere saggi,

5
00:00:25,000 --> 00:00:29,980
o esperti medici che utilizzano strumenti di intelligenza artificiale per far progredire la ricerca sul cancro.

6
00:00:30,360 --> 00:00:33,260
Ma poi ci sono anche quelle foto generate dall'intelligenza artificiale,

7
00:00:33,400 --> 00:00:35,340
video e musica che girano,

8
00:00:36,000 --> 00:00:41,580
e quelli hanno suscitato preoccupazioni sul copyright e anche sulla diffusione della disinformazione.

9
00:00:41,960 --> 00:00:42,500
Quindi legislatori,

10
00:00:43,040 --> 00:00:43,900
in particolare in Europa,

11
00:00:44,340 --> 00:00:46,440
hanno deciso di fare qualcosa al riguardo,

12
00:00:46,660 --> 00:00:49,520
introdurre la legge sull'intelligenza artificiale.

13
00:00:49,560 --> 00:00:53,800
I membri del Parlamento europeo vogliono assicurarsi che l'IA sia sicura,

14
00:00:54,320 --> 00:00:54,860
trasparente,

15
00:00:55,300 --> 00:00:55,940
tracciabile,

16
00:00:56,260 --> 00:00:59,080
non discriminatorio e rispettoso dell'ambiente.

17
00:00:59,320 --> 00:01:01,820
E se approvato, sarà il primo dei suoi

18
00:01:01,820 --> 00:01:03,140
gentile nel mondo.

19
00:01:03,580 --> 00:01:07,900
L'idea è di governare l'uso dell'IA sulla base di tre livelli di rischio,

20
00:01:08,200 --> 00:01:10,600
a partire da un rischio inaccettabile.

21
00:01:11,080 --> 00:01:14,980
Cioè, quando l'IA viene usata per cose ritenute così immorali,

22
00:01:15,100 --> 00:01:16,920
come la sorveglianza biometrica,

23
00:01:17,620 --> 00:01:21,160
o addirittura usarlo per mantenere i punteggi sociali sulle persone.

24
00:01:21,440 --> 00:01:23,120
Pensa a Black Mirror di Netflix.

25
00:01:23,600 --> 00:01:24,960
Il prossimo è ad alto rischio,

26
00:01:25,080 --> 00:01:28,300
quindi cose che potrebbero causare danni alla salute delle persone,

27
00:01:28,740 --> 00:01:29,720
danni all'ambiente,

28
00:01:29,920 --> 00:01:32,140
o ledere i diritti fondamentali delle persone.

29
00:01:32,660 --> 00:01:37,220
Ad esempio, uno strumento di intelligenza artificiale che scansiona i CV per classificare i candidati al lavoro,

30
00:01:37,260 --> 00:01:40,480
che va bene, fintanto che si attiene a certi

31
00:01:40,480 --> 00:01:40,880
regole.

32
00:01:41,000 --> 00:01:46,060
Ad esempio, non discrimina i candidati in base alla loro razza, età o sesso.

33
00:01:46,600 --> 00:01:53,120
Successivamente ci sono app e strumenti di intelligenza artificiale che non sono vietati e non presentano alcun rischio elevato.

34
00:01:53,300 --> 00:01:56,340
Quindi, quando questo atto di intelligenza artificiale diventerà legge?

35
00:01:56,580 --> 00:01:57,040
BENE,

36
00:01:57,400 --> 00:01:58,760
potrebbe volerci ancora del tempo.

37
00:01:58,940 --> 00:02:02,920
Il Parlamento europeo voterà l'attuale progetto a giugno,

38
00:02:02,920 --> 00:02:04,420
e anche allora sarà solo

39
00:02:04,420 --> 00:02:09,900
diventare legge una volta che il Parlamento europeo e il Consiglio europeo l'abbiano concordato.

40
00:02:10,480 --> 00:02:10,560
Grazie per la visione.

Spanish transcription text:

PONENTE 1: Los políticos en Europa votarán sobre una propuesta para introducir una ley que rija el uso de la inteligencia artificial. La medida se produce después de las advertencias de los expertos sobre los graves riesgos y consecuencias si continúa sin regularse. Aquí está Simi Jolla -Osso con más.

ALTAVOZ 2: La inteligencia artificial ha estado mucho en las noticias últimamente, ya sea que se trate de estudiantes que usan aplicaciones como ChatGPT para escribir ensayos o expertos médicos que usan herramientas de IA para avanzar en la investigación sobre cáncer. Pero también están esas fotos, videos y música generados por IA que circulan, y eso ha despertado preocupaciones sobre los derechos de autor y también sobre la difusión de desinformación. Entonces, los legisladores, específicamente en Europa, han decidido hacer algo al respecto, introduciendo la Ley de Inteligencia Artificial. Los miembros del Parlamento Europeo quieren asegurarse de que la IA sea segura, transparente, rastreable, no discriminatoria y respetuosa con el medio ambiente. Y si se aprueba, será el primero de su tipo en el mundo. La idea es gobernar el uso de la IA en base a tres niveles de riesgo, comenzando con el riesgo inaceptable. Es decir, cuando la IA se usa para cosas que se consideran tan poco éticas, como la vigilancia biométrica, o incluso para llevar el puntaje social de las personas. Piensa en Black Mirror de Netflix. El siguiente es alto riesgo, es decir, cosas que pueden causar daño a la salud de las personas, dañar el medio ambiente o afectar los derechos fundamentales de las personas. Por ejemplo, una herramienta de inteligencia artificial que escanea CV para clasificar a los solicitantes de empleo, lo cual está bien, siempre que cumpla con ciertas reglas. Por ejemplo, no discrimina a los solicitantes en función de su raza, edad o género. A continuación, se encuentran las aplicaciones y herramientas de IA que no están prohibidas y no representan un alto riesgo. Entonces, ¿cuándo se convertirá en ley esta ley de IA? Bueno, todavía puede tomar algún tiempo. El Parlamento Europeo votará el proyecto actual en junio, e incluso entonces solo se convertirá en ley una vez que tanto el Parlamento Europeo como el Consejo Europeo lo hayan acordado. Gracias por mirar.

Spanish transcription subtitles:

0
00:00:00,740 --> 00:00:06,440
Los políticos en Europa votarán una propuesta para introducir una ley que rija el uso de la inteligencia artificial.

1
00:00:06,860 --> 00:00:13,080
La medida se produce después de las advertencias de los expertos sobre los graves riesgos y consecuencias si continúa sin regularse.

2
00:00:13,660 --> 00:00:15,960
Aquí está Simi Jolla-Osso con más.

3
00:00:17,340 --> 00:00:20,900
La inteligencia artificial ha estado mucho en las noticias últimamente,

4
00:00:21,040 --> 00:00:24,880
ya sean estudiantes que usan aplicaciones como ChatGPT para escribir ensayos,

5
00:00:25,000 --> 00:00:29,980
o expertos médicos que utilizan herramientas de inteligencia artificial para avanzar en la investigación sobre el cáncer.

6
00:00:30,360 --> 00:00:33,260
Pero también están esas fotos generadas por IA,

7
00:00:33,400 --> 00:00:35,340
vídeos y música circulando,

8
00:00:36,000 --> 00:00:41,580
y esos han despertado preocupaciones sobre los derechos de autor y también sobre la difusión de desinformación.

9
00:00:41,960 --> 00:00:42,500
Así que los legisladores,

10
00:00:43,040 --> 00:00:43,900
específicamente en Europa,

11
00:00:44,340 --> 00:00:46,440
han decidido hacer algo al respecto,

12
00:00:46,660 --> 00:00:49,520
la introducción de la Ley de Inteligencia Artificial.

13
00:00:49,560 --> 00:00:53,800
Los miembros del Parlamento Europeo quieren asegurarse de que la IA sea segura,

14
00:00:54,320 --> 00:00:54,860
transparente,

15
00:00:55,300 --> 00:00:55,940
rastreable,

16
00:00:56,260 --> 00:00:59,080
no discriminatoria y respetuosa con el medio ambiente.

17
00:00:59,320 --> 00:01:01,820
Y si se aprueba, será el primero de sus

18
00:01:01,820 --> 00:01:03,140
amable en el mundo.

19
00:01:03,580 --> 00:01:07,900
La idea es regir el uso de la IA en base a tres niveles de riesgo,

20
00:01:08,200 --> 00:01:10,600
comenzando con un riesgo inaceptable.

21
00:01:11,080 --> 00:01:14,980
Es decir, cuando la IA se usa para cosas que se consideran tan poco éticas,

22
00:01:15,100 --> 00:01:16,920
como la vigilancia biométrica,

23
00:01:17,620 --> 00:01:21,160
o incluso usarlo para mantener puntajes sociales en las personas.

24
00:01:21,440 --> 00:01:23,120
Piensa en Black Mirror de Netflix.

25
00:01:23,600 --> 00:01:24,960
El siguiente es alto riesgo,

26
00:01:25,080 --> 00:01:28,300
por lo que las cosas que pueden causar daño a la salud de las personas,

27
00:01:28,740 --> 00:01:29,720
daños al medio ambiente,

28
00:01:29,920 --> 00:01:32,140
o afectar los derechos fundamentales de las personas.

29
00:01:32,660 --> 00:01:37,220
Por ejemplo, una herramienta de inteligencia artificial que escanea CV para clasificar a los solicitantes de empleo,

30
00:01:37,260 --> 00:01:40,480
lo cual está bien, siempre y cuando cumpla con ciertos

31
00:01:40,480 --> 00:01:40,880
normas.

32
00:01:41,000 --> 00:01:46,060
Por ejemplo, no discrimina a los solicitantes en función de su raza, edad o sexo.

33
00:01:46,600 --> 00:01:53,120
A continuación, se encuentran las aplicaciones y herramientas de IA que no están prohibidas y no representan un alto riesgo.

34
00:01:53,300 --> 00:01:56,340
Entonces, ¿cuándo se convertirá en ley esta ley de IA?

35
00:01:56,580 --> 00:01:57,040
Bien,

36
00:01:57,400 --> 00:01:58,760
todavía podría tomar algún tiempo.

37
00:01:58,940 --> 00:02:02,920
El Parlamento Europeo votará el proyecto actual en junio,

38
00:02:02,920 --> 00:02:04,420
e incluso entonces solo será

39
00:02:04,420 --> 00:02:09,900
convertirse en ley una vez que tanto el Parlamento Europeo como el Consejo Europeo lo hayan acordado.

40
00:02:10,480 --> 00:02:10,560
Gracias por ver.

START TRANSCRIBING NOW!

Top of Page